Solenoid driver

[earckens]

Hi Keepitsimplestupid, thank you for your input but relays are not exactly what I want to include.

 

[spec]

Hi earckens,

Just to get an idea of the scope of this: what would be the maximum number of coils (left and right) that would be required by a railway setup- or is that like asking how long is a piece of string.

Also, in terms of user controls, and forgetting about the electronics for the moment, what would be the ultimate control for the points. I keep harping on about leavers, like in a signal box, because I would like that, but is this what the model railway aficionados would like?

spec

PS, depending on what you say, I may have a cunning plan

 

[earckens]

Hi spec,
the answer to your question is like asking a modelrailroader which layout he likes best .

However, I will give it my subjective go:
1. manageable size: in terms of turnouts not more then 50
2. modern concept: DCC (I start using DCC++ with Arduino), minimal relays, but with a user hardware control panel (including turnout controls and turnout indicators)
3. turnout control: manual, by means of physical switches, some by DCC. Switches need to be double throw type with either "left" or "right" positioning: visual indicator of turnout position. To be complemented later with LED indicators showing either left or right turnout position on the control panel.
4. turnout drive: electrical/electronic (chemelec's ideas are very appealing). No relays. Solenoids (and at later stage servomotors) driving the "points".
5. optionally reading the turnout position (left or right) into Arduino for further processing (driving layout track signaling, possibly also for interlocking and/or automatic driving (then we start talking "block detection").

Now I am so curious about your further input . If you need any additional info please let me know.
Thanks,
Erik

 

[spec]

Thank for comprehensive reply earckens.

One question: you use the term 'turnout' is that the same as 'points' in my language?

The cunning scheme I have in mind is to uses one toggle switch per point (L&R coils) which connect to an Arduino.
The Arduino in turn feeds 8 channel driver chips, one channel being connected to a single point coil. For example 8 driver chips would handle 48 coils.

All pulse timing etc would be done in software, which would be pretty straight forward.

You could add as many switches and point coils that you liked.

All the other functions that you mentioned sound pretty straight forward once you have a microcontroller in the system (famous last words).

spec

 

[chemelec]

Thank you Chemelec. But now if multiple turnouts have to be activated I should place their respective DPDT ON-ON switches in parallel, right? Now this results in power going to one solenoid per turnout.
Would using double pole pushbuttons not be an alternative (see my drawing proposal)?

Yes the Other Switch and solenoids Parallel The one I Show.

And Using a Center OFF, DPDT Switch is the BEST Solution.
This way only one solenoid will be active and this will be Easily Visible as to which one is being active and When in Center, All are Disconnected.

Your DPST PUSHBUTTON Still requires another Switch to Determine Which Solenoid.

 

[earckens]

Your DPST PUSHBUTTON Still requires another Switch to Determine Which Solenoid.

I disagree: every SPDT switch is in series with a DPST pushbutton; if a turnout has to be thrown, then the one corresponding pushbutton is pressed, the one pole closes the circuit between the MOSFET and the SPDT switch, the other pole activates the input of the circuit to the MOSFET to enable it.
Or do I miss something?

 

[earckens]

One question: you use the term 'turnout' is that the same as 'points' in my language?

Sorry for the confusion, yes I meant points to be the same as turnout

The cunning scheme I have in mind is to uses one toggle switch per point (L&R coils) which connect to an Arduino.

one toggle (SPDT?) per turnout, so left and right coil.. which is what we more or less have now: perfect.

The Arduino in turn feeds 8 channel driver chips, one channel being connected to a single point coil. For example 8 driver chips would handle 48 coils.

8x8=64?

All pulse timing etc would be done in software, which would be pretty straight forward.

You could add as many switches and point coils that you liked.

All the other functions that you mentioned sound pretty straight forward once you have a microcontroller in the system (famous last words).

spec

I prefer the hardware solution (i.e. chemelec) for driving layout basics (turnouts, ..) and keep software for lighting ans signaling: 1. turnout control is already by means of switches 2. hardware is required anyway to control and direct power, I feel there is no direct added value in using software to drive turnouts (interlocking is another matter, but here it is simple "left" or right" per turnout.

 

[spec]

Sorry for the confusion, yes I meant points to be the same as turnout

Thanks

one toggle (SPDT?) per turnout, so left and right coil.. which is what we more or less have now: perfect.

Also, no push buttons: just throw the single pole double throw toggle switch and the points (turnouts) change.

8x8=64?

I prefer the hardware solution (i.e. chemelec) for driving layout basics (turnouts, ..) and keep software for lighting ans signaling: 1. turnout control is already by means of switches 2. hardware is required anyway to control and direct power, I feel there is no direct added value in using software to drive turnouts (interlocking is another matter, but here it is simple "left" or right" per turnout.

Fair enough.

spec

 

[earckens]

Just one SPDT toggle switch (no center position, so ON-ON, right?): now I am curious

 

[chemelec]

This is what I Recommend.
If you want to try something different, Go ahead.

 

[chemelec]

Just one SPDT toggle switch (no center position, so ON-ON, right?): now I am curious

So Than All Solenoids will use Power!

 

[spec]

Just one SPDT toggle switch (no center position, so ON-ON, right?): now I am curious

Yes that is correct.

This is what I propose:

The toggle switches are arranged in an N by N matrix and the Arduino continually scans the switch positions and keeps a look up table in memory. If the Arduino detects a change in switch position it pulses the appropriate coil.

Every so often the Arduino could step through the look up table an simply pulse each coil called to do an absolute configuration to give some belt and braces.

The Arduino would be able to set any time pulse to any degree of accuracy (within reason).

As stated the coils are pulsed via standard & cheap, driver chips which have 8 channels. The Arduino selects which channel to pulse.

I like chemelec's pulse circuit, but you either have the option of providing a pulse circuit for each coil or using some time dependent function which is fairly complex and possibly prone to glitches.

It also compromises the user interface, and has less flexibility, in my opinion.

It would then be fairly straight forward to hook up a small graphics display (cost around £6UK) to the Arduino and then the sky would be the limit.

You mention point tell backs: that would also be simple with an Arduino at the heart of the system.

I would be inclined to go for an Arduino Megga (£8UK) though, rather than a Zero.

The software would be simple to write and good fun.

 

[earckens]

I like chemelec's pulse circuit, but you either have the option of providing a pulse circuit for each coil or using some time dependent function which is fairly complex and possibly prone to glitches.

I do not agree: with the additional double pole pushbuttons (see my last uploaded circuit drawing) it is possible to power only the one coil where power is required

It also compromises the user interface, and has less flexibility, in my opinion.

Why? With the ON-ON SPDT switch there is visual display of the turnout position. And further LED indicators are possible (introdice Arduino)

It would then be fairly straight forward to hook up a small graphics display (cost around £6UK) to the Arduino and then the sky would be the limit.

Right, but for basic turnout activation I would prefer a hardware solution. Which according to my facts is possible (circuit drawing with double pole pushbuttons); unless I will be contradicted

You mention point tell backs: that would also be simple with an Arduino at the heart of the system.

Right.

 

[chemelec]

Earckens, On your circuit, you are using that DPDT Push Button to trigger it and Also another Switch "S1" to determine Left or Right.

With my DPDT, Center Off toggle switch, THAT is the Only Switch required.
Center is OFF, Flip it one way or the Other to move it Right or Left and also trigger it.

But Whatever you think is best for you.

 

[earckens]

Chemelec, I fully agree with you. But as I see it the only drawback on your proposal is that there is no visual cue on the SPDT switch: an operator cannot tell which way (left or right) the turnouts are positioned.

So either I will have to introduce another way to visualise the turnout positions (with LED's for example, on the control panel) if I go the "center-toggle-off-switch" solution, or I stick with a probably old-fashioned way of visualising the turnout position (with ON-ON SPDT switches). Now that I start to think about it, altough your solution is not necessarily cost effective (only one pushbutton, but then the higher cost of center-rest switches versus ON-ON switches will offset that), it is more elegant (less clutter on the control panel) and more modern (turnout position indication through LED...

Something to be sunk in, chew over, sleep over,.. but your solution I feel is the more modern one.

 

[earckens]

To chemelec, your post #70: supply is 15V, not 25V: what components require value adjustment? I can think of the RC combo on pin 6 & 7 only?

 

[earckens]

Now there is a drawback on your design chemelec: to activate a turnout, both the pushbutton and center-rest switch have to be pressed simultaneously...

 

[chemelec]

To chemelec, your post #70: supply is 15V, not 25V: what components require value adjustment? I can think of the RC combo on pin 6 & 7 only?

If your Supply is 15 Volts, Than the Zener and the 680 Resistor are Not Required.
I'm Sure One of your drawings showed 25 Volts to the Solenoids.

 

[chemelec]

Now there is a drawback on your design chemelec: to activate a turnout, both the pushbutton and center-rest switch have to be pressed simultaneously...

The DPDT Switch Not Only Triggers it, It determine Which Way.
Example:
If Lever Straight Up is OFF.
If Lever is Flipped Left Triggers LEFT Solenoid.
If Lever is Flipped Right Triggers RIGHT Solenoid.

ALL Very VISABLE To your Eyes!

 

[chemelec]